1. Field of the Invention
This invention relates generally to apparatus and methods for the calibration of implanted pressure transducers.
2. Description of the Related Art
Pressure transducers have been used to measure physiologic pressures in a variety of locations within the body including the major blood vessels, cardiac chambers, body cavities, viscera, and tissues. Pressure transducer types have included devices with differing mechanisms of action such as: piezoelectric crystals, optical, capacitance, inductance, electrolytic, and resistive strain gauge manometers. Manometer theory, invasive physiologic pressure measurement, recording, and calibration are discussed in Mackay, Nichols et al., and Milnor, all herein incorporated by reference.
The output of a pressure transducer is a signal that is related in a known way to the pressure. In some cases, the signal is electrical in nature, such as a resistance, capacitance, inductance, or voltage that changes as a function of applied pressure. In one aspect, the signal has many different forms, depending on the type of transducer. Some other examples of pressure transducer signals include, but are not limited to, change of the frequency of an oscillator in response to pressure, change in color in response to pressure, and change in position of an indicator dial in response to pressure. It will be clear to one skilled in the art that various embodiments of the present invention applies equally well to any form of pressure transducer output signal. In one aspect, the output signal of a pressure transducer is a one to one function of the applied pressure, and in general depends on other variables. This relationship can be written as:s=f(P,c0,c2, . . . , cn);where s is the transducer output signal, which is a function, f, of:P, the pressure applied to the transducer; andc0, c1, c2, . . . , cn, additional parameters that determine the relationship between input pressure and output signal.
To obtain a measurement of pressure from the pressure transducer signal, the inverse function of f(P), written f−1(s), must be known. This can be written as:P′=f−1(s,c0,c1,c2, . . . , cn)where P′ is the measured pressure andwhere the inverse function also depends on the additional parameters c0, c1, c2, . . . , cn.
For example, in a typical linear pressure measurement system, the system can be approximated by the expression:s=a·(P−P0)=f(P,c0,c1),where c0 and c1 are identified with a and P0, respectively;s is the transducer's output signal;P is the physiologic pressure that is detected by the pressure transducer, such as a resistive strain gauge;P0 is the pressure at which the transducer output is zero, sometimes called the “baseline pressure;” anda is the “gain.”
In this simple example, it is clear that the transducer signal is a linear function of pressure with a equal to the slope of the signal versus pressure plot, and P0 and −a P0 equal the x and y intercepts, respectively.
The inverse function is obtained by solving the output function for the pressure. In this example:P′=(1/α′)·s+P0′where P′ is the measurement of the pressure;a′ is the assumed value of the gain parameter. The quantity (1/a′) is sometimes called the “scale factor” or “calibration factor;”s is the transducer output signal; andP0′ is the assumed value of the baseline pressure, also known as the “offset.”
If the assumed values for the parameters, a′ and P0′, are equal to the true values, a and P0, that determined the output of the transducer, the measured value P′ will be equal to the true pressure P, and the transducer is said to be calibrated.
Thus, calibration of the transducer consists in general of determining the values of the parameters c0, c1, c2, . . . , cn of the transducer output function, in this example the two parameters a and P0 (gain and offset), so that the inverse function of the transducer output signal will be equal to the true pressure. One skilled in the art will understand that although the transducer output function in this example was linear, the discussion of calibration is not limited to linear functions. However, the output signal should preferably be a one-to-one function of applied pressure, which guarantees that the inverse function exists and is also a one-to-one function, allowing the pressure to be calculated from the output signal once the transducer is calibrated. In some cases, a pressure transducer is calibrated by applying known pressures, observing the corresponding output signals, then solving the system of equations relating the known pressures and observed outputs for the unknown parameter values. In general, the minimum number of different pressure-output pairs required to solve for the parameters is equal to the number of unknown parameters. In the above example, there were two parameters, gain and offset, so at least two different pressures would need to be applied and the output signal recorded to determine both parameters.
Calibration may be performed at the time of manufacture of the transducer system. The transducer will then accurately measure true pressure as long as the true parameters determining transducer output remain constant. In reality, however, parameters such as gain and offset may change over time, a phenomenon known as “drift.” Indeed, currently available transducer devices experience some degree of drift. Because of drift, transducer devices must be periodically recalibrated to ensure accurate readings.
A number of factors may contribute to this drift effect. These factors include changing atmospheric pressure, temperature, humidity, damping, material creep, fatigue, and aging of pressure transducer and electronic components. For example, the offset of a transducer incorporating a sealed chamber with an internal pressure that defines the baseline pressure will drift if the sealed chamber leaks so that its internal pressure changes. In this example, the direction of drift is determined by the initial pressure differential between the internal and external pressure. If the external pressure is lower than the internal pressure, the leak will cause an upward drift in pressure readings. If the internal pressure is lower, the leak will cause a downward drift in pressure readings. The rate of a leak is generally directly proportional to the pressure differential. In one aspect of the present invention, the rate of leak is used to maintain calibration over time.
In another example, drift due to material creep, or viscoelastic behavior, is a contributing factor in pressure transducer drift. This effect may also produce predictable calibration changes over time that depend on the intrinsic viscoelastic properties of the transducer, the pressure differential, the temperature, and even the past history of the pressure differential. In one aspect of the present invention, measured drift characteristics of a pressure transducer are utilized together with its temperature and pressure history to predict and correct for transducer drift.
As described below, various methods are known for recalibrating physiologic pressure transducers in clinical use. Each of these requires some means of access to the transducer for applying known pressures. An object of the present invention is to provide a method for recalibrating a pressure transducer that is implanted within the patient's thoracic cavity where previous methods for applying known pressure for calibration are either too invasive or do not work.
One common method to measure physiologic pressure uses a strain gauge type transducer located external to the body with a sensing membrane that is displaced by a first side being in continuity to a fluid filled catheter that communicates with the location where pressure measurement is desired. Physiologic pressure is typically measured as gauge pressure, which represents the differential of the absolute internal pressure and ambient air pressure. This is accomplished by having the second side of the sensing membrane in communication with air so that the membrane moves in response to the differential pressure. Such transducers are readily recalibrated. To do so, the first side of the membrane is temporarily exposed to air. The offset parameter is then adjusted until the pressure reading is zero. Next, the first side of the membrane is temporarily exposed to a known pressure, classically a vertical column of mercury, while the gain parameter is adjusted to match the known pressure head provided by the mercury column. Although this recalibration technique is commonly used, other physiologic pressure measuring and calibration methods are known to those skilled in the art.
To achieve higher fidelity with physiologic pressure signals, transducers have been placed in the body by mounting the transducers on or near the distal tip of a diagnostic catheter. Hamatake in U.S. Pat. No. 5,788,642 describes an apparatus for re-zeroing a catheter-based pressure transducer when it is at the in vivo measurement site in the body. This apparatus provides for a means of exposing both sides of the pressure transducer either to atmospheric pressure or to physiologic pressure to re-zero the pressure transducer.
In another approach, Demarest in U.S. Pat. No. 4,886,070 describes an apparatus for recalibrating both offset (zero) and gain of a catheter-based pressure transducer while it is at the in vivo measurement site in the body. This apparatus provides for a lumen within the catheter in communication at the distal end with the inside of the “pressure responsive element” (e.g., a diaphragm) and which is accessible at the proximal end outside the patient. In this invention, the pressure responsive element presses against the strain gauge via, e.g., a strut attached either to the diaphragm or the strain gauge, but not both. The system is manufactured so that when both the outside (measurement side) and the inside of the diaphragm are at the same pressure, the strut presses against the strain gauge, so that the strain gauge is said to be preloaded. An increase in measured pressure will further strain the gauge. The measured pressure signal is the change in resistance between the preloaded strain and the additional strain. The calibration method consists of increasing the pressure on the inside via the lumen until the indicated strain ceases to decrease, corresponding to the point at which the strut no longer presses against the strain gauge and the gauge is unloaded. One disadvantage of this approach is that it assumes that the “zero reference back pressure” (e.g., the inside pressure required to just unload the strain gauge) never drifts. This assumption does not necessarily hold, due to, for example, aging and creep in the materials and adhesives used to construct both the pressure responsive element, the coupling strut, and the strain gauge. In addition, such a device can only be used temporarily because recalibration requires access via a catheter through the patient's skin in order to manipulate the pressure inside the transducer.
Trimble, in U.S. Pat. No. 5,437,284, describes an essentially similar apparatus and method. In Trimble, however, a mechanical limit is used to establish a reference position of the pressure responsive element, instead of the point of decoupling of the pressure responsive element and the strain gauge, as taught by Demarest. As with Demarest, Trimble is based upon the possibly flawed assumption that this reference position does not itself drift and, like Demarest, Trimble requires application of a known pressure to the inside of the transducer, requiring access to the transducer through the patient's skin.
Thus, calibration issues have relegated catheter mounted pressure transducers to very limited application for patient monitoring. Indeed, use of these devices has been restricted to research studies in human patients for up to a few days or for up to several weeks in laboratory animals.
In recent years there has been a growing interest in implantable pressure transducers that can be used to diagnose and guide therapy in medical patients. Checking and maintaining calibration for such chronically implanted transducers is especially problematic because the transducer cannot easily be directly accessed to provide zeroing and reference pressures. Measuring gauge pressure requires a transducer scheme using one or two transducers. In the two-transducer scheme, the first transducer measures absolute pressure at the desired location and the second transducer, which measures absolute atmospheric pressure, is subtracted from the first. A single transducer scheme requires that a transducer has a first side of its sensing unit (diaphragm or membrane) exposed to the location of the desired pressure measurement and its second side exposed to the ambient atmosphere or its equivalent. Having direct continuity to the ambient atmosphere may not be practicable because this creates a path for the ingress of infective organisms. Several calibration methods for implanted pressure transducers have been described. Attempts have been made to use the interstitial pressure in the subcutaneous spaces as an atmospheric reference equivalent. Subcutaneous pressure, however, may differ from atmospheric pressure for a variety of reasons, especially when there are rapid changes in atmospheric pressure.
Meador, in U.S. Pat. No. 6,234,973, describes a pressure monitor for a cardiac pacemaker where a first transducer is used to measure a physiologic pressure and a second transducer is used to supply an atmospheric reference pressure. The second transducer is located superficially at or near the pacemaker generator to provide compensation for changes in atmospheric pressure. This transducer can be located subcutaneously with a subcutaneous access port that can be entered with a hypodermic needle for calibration of pressure. Such an arrangement would allow for calibration of the second subcutaneous transducer, but has the disadvantage of requiring penetration of the skin with the attendant discomfort and risk of infection. The device described in Meador has the additional limitation that the primary transducer located in the heart is not calibrated.
Cosman, in U.S. Pat. Nos. 4,676,255 and 4,206,761, describes a calibration method for an intracranial implantable pressure sensor that does not require direct access to the transducer via catheter or hypodermic needle. Rather, in vivo calibration is performed utilizing variations of positive and negative pressures applied via a chamber sealed against the skin overlying a single transducer. The sensor in this case was mounted through the skull with the inner side (first side) of the pressure responsive element exposed to the intracranial pressure and the outer side (second side) in contact with the scalp. This approach still has the disadvantage that a mechanical stop must be provided such that the differential pressure needed to drive the pressure responsive element to this reference position never changes over time. Thus, it can provide a calibration for gain, but not necessarily for offset. Despite this disadvantage, the method described does provide a means for in vivo calibration of gain (and offset if the assumption holds true) without direct access to the transducer through the skin. However, the method of using a chamber against the skin overlying the transducer would work only for those cases where access to the second side the transducer is located very superficially.
Although an implanted transducer can be calibrated just prior to implantation, even without removal from sterile packaging (as described in U.S. Pat. No. 6,292,697, incorporated by reference herein), there is no assurance that the transducer will remain calibrated. Further, once implanted, calibration cannot be easily verified without performing an invasive procedure, such as insertion of a second calibrated transducer into the body, positioned in a location where the pressure is sufficiently similar to that at the location of the permanent transducer. Such an invasive procedure may have associated risks to the health of the patient. In one study in which this was performed, Magalski et al., reported on an implanted pressure-measuring device that uses an algorithm applied to the right ventricular pressure tracing to estimate pulmonary artery diastolic pressure (PADP). PADP is a well-established surrogate for estimating the left atrial pressure (LAP), which is one of the key predictors of worsening heart failure. Initial calibrated baseline recordings in 32 patients with heart failure showed that the estimated PADP reading differed on average from the true, invasively obtained PADP by only −0.1±5.5 mm Hg. However, one year after implantation, invasive recalibration showed significant drift with an average underestimation of PADP by −3.6±6.9 mm Hg. Ultimately, transducer drift may be so significant that the measurement data become clinically useless without adequate recalibration.
Difficulty with maintaining pressure transducer calibration raises the issue of how much miscalibration is acceptable for diagnostic accuracy, especially when the results are used to make therapeutic decisions. This is important because pressure variations as small as 5 mm Hg may alter therapeutic interventions. For example, a patient with congestive heart failure will often be clinically stable and feel well (known as a condition of “compensated” heart failure) with an elevated left atrial pressure of 20 mm Hg. Such a patient may start to “decompensate,” with fluid beginning to enter the lungs eventually resulting in clinical symptoms such as shortness of breath, when the left atrial pressure increases to 25 mm Hg. A patient being managed using left atrial pressure to catch the early onset of decompensation would be treated to reduce LAP, either by changing oral medications, administering drugs by injection, or even automatically delivering drugs, electrical pacing, or other therapy by an implanted device based on pressure measurements. Consequently, calibration drift errors like those reported above may be large enough to inappropriately influence medical treatment of the patient. The fact that calibration drift can profoundly affect a patient's diagnosis and ensuing therapy underscores the importance of periodic recalibration, which currently requires an invasive procedure to place a second calibrated transducer in a suitable location to assure an accurate comparison.
As discussed above, currently used methods of monitoring and maintaining calibration of implanted pressure transducers possess significant drawbacks. In various embodiments of the present invention, the calibration of implanted pressure transducers can be routinely checked and, whenever necessary, recalibrated using less-invasive methods and apparatus than those currently available. These advantages, among others, will be further understood and appreciated by reference to the written disclosure, figures, and claims herein.